An ensemble-based method for linear feature extraction for two-class problems

In this paper we propose three variants of a linear feature extraction technique based on Adaboost for two-class classification problems. Unlike other feature extraction techniques, we do not make any assumptions about the distribution of the data. At each boosting step we select from a pool of linear projections the one that minimizes the weighted error. We propose three different variants of the feature extraction algorithm, depending on the way the pool of individual projections is constructed. Using nine real and two artificial data sets of different original dimensionality and sample size we compare the performance of the three proposed techniques with three classical techniques for linear feature extraction: Fisher linear discriminant analysis (FLD), Nonparametric discriminant analysis (NDA) and a recently proposed feature extraction method for heteroscedastic data based on the Chernoff criterion. Our results show that for data sets of relatively low-original dimensionality FLD appears to be both the most accurate and the most economical feature extraction method (giving just one-dimension in the case of two classes). The techniques based on Adaboost fare better than the classical techniques for data sets of large original dimensionality.

---

A SIMD-efficient 14 instruction shader program for high-throughput microtriangle rasterization

This paper shows that breaking the barrier of 1 triangle/clock rasterization rate for microtriangles in modern GPU architectures in an efficient way is possible. The fixed throughput of the special purpose culling and triangle setup stages of the classic pipeline limits the GPU scalability to rasterize many triangles in parallel when these cover very few pixels. In contrast, the shader core counts and increasing GFLOPs in modern GPUs clearly suggests parallelizing this computation entirely across multiple shader threads, making use of the powerful wide-ALU instructions. In this paper, we present a very efficient SIMD-like rasterization code targeted at very small triangles that scales very well with the number of shader cores and has higher performance than traditional edge equation based algorithms. We have extended the ATTILA GPU shader ISA (del Barrioet al. in IEEE International Symposium on Performance Analysis of Systems and Software, pp. 231–241, 2006) with two fixed point instructions to meet the rasterization precision requirement. This paper also introduces a novel subpixel Bounding Box size optimization that adjusts the bounds much more finely, which is critical for small triangles, and doubles the 2×2-pixel stamp test efficiency. The proposed shader rasterization program can run on top of the original pixel shader program in such a way that selected fragments are rasterized, attribute interpolated and pixel shaded in the same pass. Our results show that our technique yields better performance than a classic rasterizer at 8 or more shader cores, with speedups as high as 4× for 16 shader cores.     

SMOS first data analysis for sea surface salinity determination

Soil Moisture and Ocean Salinity (SMOS), launched on 2 November 2009, is the first satellite mission addressing sea surface salinity (SSS) measurement from space. Its unique payload is the Microwave Imaging Radiometer using Aperture Synthesis (MIRAS), a new two-dimensional interferometer designed by the European Space Agency (ESA) and operating at the L-band frequency. This article presents a summary of SSS retrieval from SMOS observations and shows initial results obtained one year after launch. These results are encouraging, but also indicate that further improvements at various data processing levels are needed and hence are currently under investigation.     

Evolutionary Multiobjective Optimization algorithm for multimedia delivery in critical applications through Content-Aware Networks

Critical applications which need to deliver multimedia through the Internet, may achieve the required quality of service thanks to the Content-Aware Networks (CAN). The key element of CAN is an efficient decision algorithm responsible for the selection of the best content source and routing paths for content delivery. This paper proposes a two-phase decision algorithm, exploiting the Evolutionary Multiobjective Optimization (EMO) approach. It allows to consider valid information in different time scales, adapting decision-maker to the evolving network and server conditions as well as to get the optimal solution in different shapes of Pareto front. The simulation experiments performed in a large-scale network model, confirm the effectiveness of the proposed two-phase EMO algorithm, comparing to other multi-criteria decision algorithms used in CAN.     

Pattern codification strategies in structured light systems

Coded structured light is considered one of the most reliable techniques for recovering the surface of objects. This technique is based on projecting a light pattern and viewing the illuminated scene from one or more points of view. Since the pattern is coded, correspondences between image points and points of the projected pattern can be easily found. The decoded points can be triangulated and 3D information is obtained. We present an overview of the existing techniques, as well as a new and definitive classification of patterns for structured light sensors. We have implemented a set of representative techniques in this field and present some comparative results. The advantages and constraints of the different patterns are also discussed.     

Purification and characterization of X-prolyl-dipeptidyl-aminopeptidase from Lactobacillus lactis and from Streptococcus thermophilus

X-Prolyl-dipeptidyl-aminopeptidase recently was found in several lactic acid bacteria. This article describes the purification of the enzymes from Lactobacillus lactis and Streptococcus thermophilus and compares their characteristics. Enzymes from both strains are serine-peptidases. They both have a molecular weight of about 165,000 daltons, an isoelectric point near 4.5, and are constituted of two subunits. The pH optimum of the enzyme isolated from L. lactis is 7.0, whereas the enzyme from S. thermophilus possesses a broad pH optimum between 6.5 and 8.2 with glycyl-L-prolyl-aminomethylcoumarin as substrate. Below pH 5, both enzymes are unstable; however, that from S. thermophilus is more rapidly denatured. The enzyme from S. thermophilus is more sensitive to heat than the corresponding enzyme from L. lactis. Enzymes from the both strains have different specificities towards various substrates and are differently effected by metals, chelators, and other inhibitors. The importance of this enzyme for the metabolism of lactic acid bacteria is discussed.     

High level communication functionalities for wireless sensor networks

In this paper we show how to establish a reliable and efficient high level communication system in a randomly deployed network of sensors equipped with directional antennas. This high level communication system enables the programming of the sensor network using high level communication functionalities without the burden of taking care of their physical capacities (low range, unidirectional links, single frequency, presence of collisions, etc.). The high level communication functionalities we offer include point-to-point communication, point-to-area communication, and one-to-all communication. The basic idea to implement this system is to simulate a virtual network that emerges from the ad-hoc network using self-organization, self-discovery and collaborative methods. We also analyse the efficiency, scalability and robustness of the proposed protocols.     

Variation of the chemical composition of tomato cultivars (Lycopersicon esculentum Mill.) according to resistance against the tomato yellow leaf curl virus (TYLCV)

BACKGROUND: Tomato producers are looking for alternatives to protect their crops against the tomato yellow leaf curl virus (TYLCV), including replacement of cultivars with other more resistant ones. Major chemical components, sugars, mineral composition, organic acids, lycopene, total phenols and hydroxycinnamic acids were determined in six tomato cultivars, three of them (Boludo, Dorothy and Tyna) resistant, and the other three (Daniela, Dominique and Thomas), non‐resistant against TYLCV. RESULTS: The Daniela cultivar showed the greatest difference with respect to the others, mostly due to the higher content of soluble solids. The major significant differences between the mean values according to the cultivar and resistance against the TYLCV were observed for total soluble solids, pH, ascorbic acid, total phenols and hydroxycinnamic acids. High coefficients of correlation were found between glucose and fructose, and among ferulic, caffeic and p‐coumaric acids. CONCLUSION: Linear discriminant analysis made it possible to differentiate tomato samples according to the cultivar. A clear varietal influence on the chemical composition was observed; however, resistance or non‐resistance against TYLCV does not seem to influence the chemical composition. Copyright © 2008 Society of Chemical Industry     

On the accuracy of the diffusion coefficient measurements using different initial shear cell configurations at low and moderate Rayleigh numbers

Three different shear cell arrangements are considered here in order to discuss and compare the accuracy of their measurements at low and moderate Rayleigh number, that is to say, in reduced gravity environments. To do so, specific numerical procedures have been designed in order to quantitatively estimate the corresponding associated errors as a function of time and at the end of the diffusive process. The present work clearly shows that the interdiffusion-couple configuration is, from the point of view of accuracy, the worst one.